Turning a folder of photos into a single PDF sounds like the most trivial conversion there is. It is also where quality quietly disappears, because the obvious way to implement it is the wrong one.
The re-encoding tax
A JPEG is not a picture. It is a compressed description of a picture, and the compression is lossy — some detail was discarded when the file was made, and it is gone.
Now consider what most converters do: decode the JPEG into raw pixels, draw those pixels onto a PDF page, and compress the result as a new JPEG. That second compression throws away detail again, on an image that has already paid the price once. Do it twice more — convert, edit, re-export — and the artefacts become visible: mushy edges, blocky skies, halos around text in a screenshot.
None of it is necessary. PDF’s native image compression is DCT, which is JPEG. A JPEG’s compressed bytes can be dropped into a PDF exactly as they are, and every reader will display them correctly. The image in the document is then bit-for-bit the image on your disk.
JPG to PDF does this for JPEGs: each one is copied into the document byte for byte, so what you put in is what comes out. A PNG is not copied that way: it is decoded and its pixels are stored again with lossless compression (8 bits per channel, without any embedded colour profile), so the PDF can come out larger than the PNG was. If your images are screenshots or graphics rather than photos, PNG to PDF covers what is different about that case — lossless storage, transparency, and why the file ends up larger. WebP to PDF covers the format you get when you save a picture from a website, which is the one format here that has to be converted to JPEG.
Carrying a JPEG in whole has one side effect: its hidden camera data goes into the PDF too. That is the date, the phone model and, if location was on when the photo was taken, GPS coordinates. A PDF viewer does not show it, but anyone who extracts the images can read it.
The other thing tools get wrong: page size
A pixel has no physical size. A 1200×800 photo is 1200 pixels wide — but how wide is that on paper? It depends entirely on the resolution you decide to print it at.
PDF measures in points, where 72 points make an inch. A converter that naively treats one pixel as one point turns that 1200-pixel photo into a page 16.7 inches — about 42 cm — across. Open it and every page is enormous; print it and nothing fits.
The right answer is to pick a resolution. Filozy reads images at 96 DPI, the long-standing screen standard, so 1200 pixels becomes 12.5 inches — roughly 32 cm — and a fitted page is never larger than A2, about 59 cm on its long side. A 1200-pixel image lands close to a normal page; a 12-megapixel phone photo is scaled down to that A2 limit, so choose A4 or Letter if it will be printed.
That is also why pages in one PDF can come out different sizes. With Fit each page to its image, the default, every page takes the shape of its own photo, so a portrait phone photo, a landscape one and a tall screenshot make three different pages. For pages that all match, choose A4 or Letter instead: each image is scaled to fit inside the same page and centred, keeping its proportions and never cropped.
Step by step
- Name the images in order first. Pages follow the order the files arrive in, and names like
01-…,02-…make a misplaced one easy to spot. - Open JPG to PDF and select them all at once. They are read in your browser; nothing is uploaded. iPhone HEIC photos are skipped, and the page says how many were left out: export them as JPEG first, or go to Settings, then Camera, then Formats and choose Most Compatible for new photos.
- Check the order. With two or more images, open Selected files above the button: it lists them in the order they will become pages. There is no drag-to-reorder on this page, so if one is in the wrong place, build the PDF anyway and move that page in Organize PDF pages.
- Select Build the PDF, then save the new PDF.
- If the file is larger than you need — a stack of phone photos easily runs to 30 MB — pass it to Compress PDF rather than shrinking the images beforehand. That way you keep the originals untouched and decide the quality trade-off once, with the real byte count in front of you.
When a photo is a document
Photographing paperwork is not quite the same job. A photo of a page has perspective distortion, uneven lighting and a background — and converting it directly gives you a PDF of a photo of a document, not a document.
Scan to PDF is built for that case: it straightens the page, corrects the lighting and crops to the sheet, all on the device. Then, if you need the text to be searchable, run OCR PDF over the result.
What to check in the finished file
- That nothing was re-compressed. Run the PDF through Extract images from PDF and compare each extracted JPEG with your original: the same size in bytes, or the same checksum, means it went in untouched. A CMYK JPEG comes out as a PNG instead, so it cannot be compared this way. A rougher sign: a PDF of JPEGs that went in untouched is about as large as the photos added together, so a PDF much larger or much smaller than that was probably re-encoded or resized.
- Orientation. Phones often save a photo sideways with a tag saying which way to turn it. JPG to PDF reads that tag on a JPEG and turns the page to match, without touching the image bytes, and the result notes say how many were turned. A PNG is placed as stored, and a photo with a mirrored tag is placed unmirrored, so open the PDF, confirm nothing arrived sideways, and turn any page that did with Rotate PDF.
- Page order, especially if the images came from more than one folder.
- File size, before you email it. A ten-photo PDF at full resolution can exceed most attachment limits.
- Legibility, if any page is a photographed document — read the smallest text you actually need at normal zoom, not zoomed in.
- Location data. A JPEG’s hidden camera data, including GPS coordinates if location was on, travels inside the PDF. If it is going to people you do not know, remove location from the photos before adding them.